The speed of processing requests is an important optimization target of data processing systems. However, the processing speed may vary based on how or where the data is being stored. For example, data can be stored in a variety of different locations in a given system, including in memory databases or disk storage databases. Data stored in memory may be retrieved and processed faster as compared to the data stored on disk, but memory based storage solutions may be some order of magnitude expensive when compared to disk based solutions. Processing requests for data that is stored on disk storage involves retrieving the data from the disk storage, processing the data in memory, and then returning the result to the requester. However, performing data processing in such a manner consumes the valuable processing capabilities of the in memory database and forces other requests to wait while the data is retrieved and processed, creating a backlog and reducing both the overall processing speed and the capabilities of the entire system.